Cherry picking and classifying method and device based on machine vision

Abstract

The present disclosure provides a cherry picking and classifying method based on machine vision, including: acquiring an image for a cherry with a stem hung on a branch, filtering the image to remove noise, and taking a resulting image as an original image; processing the original image to extract a stem feature of the cherry, and determining a picking point for picking; acquiring an image for a picked cherry to obtain a classification image; processing the classification image to extract a contour feature of the cherry, and calculating a size of the cherry according to the contour feature; and classifying the cherry according to size data of the cherry. The present disclosure further provides a picking and classifying device, including: an acquisition module, a classifying transmission module, a connection module; and a control module.

What is claimed is: 
     
         1 . A cherry picking and classifying method based on machine vision, comprising:
 acquiring an image for a cherry with a stem hung on a branch, filtering the image to remove noise, and taking a resulting image as an original image;   processing the original image to extract a stem feature of the cherry, and determining a picking point for picking, comprising:
 performing grayscale extraction on the original image with a color factor to obtain an original grayscale image; 
 transforming the original image into a hue-intensity-saturation (HIS) model image, and performing H-based segmentation to obtain a mask image; 
 excluding a fruit region in combination with the original grayscale image and the mask image to obtain a grayscale image for the stem and a neighboring region thereof; 
 binarizing the grayscale image with an Otsu algorithm to obtain a connecting point between the stem and the fruit, and a growing point between the stem and the branch; and 
 performing weighted calculation on the connecting point and the growing point to obtain the picking point; 
   acquiring an image for a picked cherry to obtain a classification image;   processing the classification image to extract a contour feature of the cherry, and calculating a size of the cherry according to the contour feature; and   classifying the cherry according to size data of the cherry.   
     
     
         2 . The cherry picking and classifying method based on machine vision according to  claim 1 , wherein the color factor is one of red (R), green (G), and blue (B) components in three primary colors or one of composite components of the three primary colors; a plurality of historically acquired cherry images are sampled; a component value of a color in a cherry region and a component value of a color in a background region are calculated with a same color factor to form a scatter plot; and a color factor making an intersection between the component values of the colors of the two regions in the scatter plot least is selected to perform the grayscale extraction on the original image. 
     
     
         3 . The cherry picking and classifying method based on machine vision according to  claim 1 , wherein a stem-branch contour is obtained from a binary image with a contour tracing algorithm; after the binary image coincides with the mask image, the stem-branch contour is intersected with a boundary of the fruit region on the mask image at a point A 1  and a point B 1 ; and a midpoint on a connecting line between the point A 1  and the point B 1  serves as the connecting point; and
 in gradual movement from the point A 1  along the stem-branch contour and away from the fruit region, there comes a point A i  that makes α=S CAi /S Ai  greater than a preset threshold; in gradual movement from the point B 1  along the stem-branch contour and away from the fruit region, there comes a point B j  that makes α=S CBj /S Bj  greater than the preset threshold; and a midpoint on a connecting line between the point A i  and the point B j  serves as the growing point, wherein   S Ai  and S Bj  are respectively a number of pixels in a circle A i  and a circle B j ; the circle A i  and the circle B j  respectively take the point A i  and the point B j  as a center and a length of a segment |A 1 B 1 | as a radius; and S CAi  and S CBj  are respectively a number of pixels within an enclosing range of the two circles and the stem-branch contour.   
     
     
         4 . The cherry picking and classifying method based on machine vision according to  claim 1 , wherein the second processing comprises:
 filtering the classification image to remove noise, and transforming a resulting image into a classification grayscale image;   processing the classification grayscale image with an edge detection operator to extract a contour of the cherry; and   calculating a number of pixels in the contour of the cherry.   
     
     
         5 . A cherry picking and classifying device based on machine vision, applied to the cherry picking and classifying method according to  claim 1 , and comprising:
 an acquisition module configured to acquire an image for a cherry with a stem hung on a branch, and pick the cherry;   a classifying transmission module configured to acquire an image for a picked cherry and calculate a size of the cherry for classifying;   a connection module configured to connect the acquisition module and the classifying transmission module, and drive the acquisition module to move; and   a control module configured to control work of the cherry picking and classifying device.   
     
     
         6 . The cherry picking and classifying device based on machine vision according to  claim 5 , wherein the acquisition module comprises at least three connecting arms; a range positioning device is provided on each of the three connecting arms; a tail end of a first connecting arm is provided with a picking device; a tail end of a second connecting arm is provided with an image acquisition device; and a tail end of a third connecting arm is provided with a placement device. 
     
     
         7 . The cherry picking and classifying device based on machine vision according to  claim 5 , wherein the classifying transmission module comprises an input unit; the input unit transmits the picked cherry to each of a plurality of branch classification units through a sorting unit; and an image acquisition unit is provided fittingly above each of the branch classification units. 
     
     
         8 . The cherry picking and classifying device based on machine vision according to  claim 5 , wherein the color factor is one of red (R), green (G), and blue (B) components in three primary colors or one of composite components of the three primary colors; a plurality of historically acquired cherry images are sampled; a component value of a color in a cherry region and a component value of a color in a background region are calculated with a same color factor to form a scatter plot; and a color factor making an intersection between the component values of the colors of the two regions in the scatter plot least is selected to perform the grayscale extraction on the original image. 
     
     
         9 . The cherry picking and classifying device based on machine vision according to  claim 8 , wherein the classifying transmission module comprises an input unit; the input unit transmits the picked cherry to each of a plurality of branch classification units through a sorting unit; and an image acquisition unit is provided fittingly above the branch classification unit. 
     
     
         10 . The cherry picking and classifying device based on machine vision according to  claim 6 , wherein a stem-branch contour is obtained from a binary image with a contour tracing algorithm; after the binary image coincides with the mask image, the stem-branch contour is intersected with a boundary of the fruit region on the mask image at a point A 1  and a point B 1 ; and a midpoint on a connecting line between the point A 1  and the point B 1  serves as the connecting point; and
 in gradual movement from the point A 1  along the stem-branch contour and away from the fruit region, there comes a point Ai that makes α=S CAi /S Ai  greater than a preset threshold; in gradual movement from the point B 1  along the stem-branch contour and away from the fruit region, there comes a point B j  that makes α=S CBj /S Bj  greater than the preset threshold; and a midpoint on a connecting line between the point A i  and the point B j  serves as the growing point, wherein   S Ai  and S Bj  are respectively a number of pixels in a circle A i  and a circle B j ; the circle A i  and the circle B j  respectively take the point A i  and the point B j  as a center and a length of a segment |A 1 B 1 | as a radius; and S CAi  and S CBj  are respectively a number of pixels within an enclosing range of the two circles and the stem-branch contour.   
     
     
         11 . The cherry picking and classifying device based on machine vision according to  claim 10 , wherein the classifying transmission module comprises an input unit; the input unit transmits the picked cherry to each of a plurality of branch classification units through a sorting unit; and an image acquisition unit is provided fittingly above the branch classification unit. 
     
     
         12 . The cherry picking and classifying device based on machine vision according to  claim 5 , wherein the second processing comprises:
 filtering the classification image to remove noise, and transforming a resulting image into a classification grayscale image;   processing the classification grayscale image with an edge detection operator to extract a contour of the cherry; and   calculating a number of pixels in the contour of the cherry.   
     
     
         13 . The cherry picking and classifying device based on machine vision according to  claim 12 , wherein the classifying transmission module comprises an input unit; the input unit transmits the picked cherry to each of a plurality of branch classification units through a sorting unit; and an image acquisition unit is provided fittingly above the branch classification unit.